Research map / Learning theory and algorithms / Matrix and tensor methods
Total Least Squares Regression in Input Sparsity Time
Huaian Diao, Zhao Song, David Woodruff, Xin Yang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use FTLS at 0.1 sample density instead of exact TLS (ablation) | FTLS · Insurance Company Benchmark | running time (s): 125.38 -> 4.894 | Improved |
| use FTLS at 0.1 sample density instead of exact TLS (ablation) | FTLS · Wine Quality White | running time (s): 29.44 -> 0.60 | Improved |
| use FTLS at 0.1 sample density instead of exact TLS (ablation) | FTLS · Airfoil Self-Noise | running time (s): 1.12 -> 0.034 | Improved |
| reduce FTLS sample density from 0.9 to 0.1 (ablation) | FTLS · Insurance Company Benchmark | regression cost: 0.553 -> 0.732 | Hurt |
| reduce FTLS sample density from 0.9 to 0.1 (ablation) | FTLS · Wine Quality Red | regression cost: 0.93 -> 0.99 | Hurt |
| reduce FTLS sample density from 0.9 to 0.1 (ablation) | FTLS · Airfoil Self-Noise | regression cost: 0.10 -> 0.10 | No clear effect |
| reduce FTLS sample density from 0.9 to 0.1 (ablation) | FTLS · Wine Quality White | regression cost: 1.857 -> 1.885 | Hurt |
| use FTLS instead of ordinary least squares (ablation) | FTLS · Insurance Company Benchmark | regression cost: 303 -> 0.553 | Improved |
| use FTLS instead of ordinary least squares (ablation) | FTLS · Airfoil Self-Noise | regression cost: $10^6$ -> 0.10 | Improved |
| use total least squares instead of least squares on toy example (ablation) | TLS · 3x2 toy matrix | cost: 9 -> 1 | Improved |
| use FTLS instead of least squares on toy example (ablation) | FTLS · 3x2 toy matrix | cost: 9 -> 1.55 | Improved |
| use FTLS instead of LS on Gaussian matrices (ablation) | FTLS · Gaussian 20k x 2k matrices | running time | Hurt |
Closest papers by meaning
- Exploiting Numerical Sparsity for Efficient Learning : Faster Eigenvector Computation and Regression (NeurIPS 2018 · similarity 0.85)
- Harder, Better, Faster, Stronger Convergence Rates for Least-Squares Regression (JMLR 2017 · similarity 0.84)
- Fast and Accurate Least-Mean-Squares Solvers (NeurIPS 2019 · similarity 0.84)
- Fast and Accurate Least-Mean-Squares Solvers for High Dimensional Data (TPAMI 2022 · similarity 0.84)
- Sublinear Time Numerical Linear Algebra for Structured Matrices (AAAI 2019 · similarity 0.84)
- Robust Sparse Regression with Non-Isotropic Designs (NeurIPS 2024 · similarity 0.81)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
